Posted by: Josh Lehner | October 19, 2016

Poverty and Progress, Portland Edition

This morning I am presenting at Multifamily NW’s Fall 2016 Apartment Report event. While the report itself shows some progress being made in terms of bringing more balance to the housing market (increased construction and rising vacancy rates), I wanted to share my portion which does pertain to the Portland metro economy. I will follow-up next week with some additional thoughts on other regions of the state.

Statewide, household incomes are rising for those in the middle and lower parts of the distribution. Like the state overall, median household income in Portland is one percent below its previous peak after taking into account inflation. Incomes in the typical large metro across the nation, however, remain some six percent below. Overall, Portland median household income is now more than 15 percent ($8,500) above the typical large metro’s income. These improvements are due to the tightening labor market.


Given that wages and the safety net make up the majority of income for such households, the strong labor market and wage gains are pulling workers and families out of poverty today. The business cycle is not dead. These improvements are seen among households in deep poverty, those with incomes less than half the official poverty level. Furthermore, poverty has improved proportionately for both whites and people of color alike. There is still a large racial gap locally and nationally. However that gap is not widening, which is better than the alternative. All told, the various poverty rates in the Portland region are now about three-fourths, or 75% recovered relative to pre-Great Recession rates.


Finally there is some good news in terms of the low-income, cost-burdened households, although it is certainly more of a mixed bag. The number of such households in the $30,000 or $40,000 range continue to be larger, even as the number making less than $20,000 is declining. Some of this is due to more jobs and rising wages, which is obviously good. However some of this improvement is likely due to inflation and some is likely due to displacement. And this is metro level displacement, where households move out of the region entirely. It is likely that all three factors are at play here, however it will require a lot more data work to try and figure out which factor is largest.


All told, the tightening labor market is beginning to create improvements in some of these deeper measures of economic well-being. As stated the other day, progress made in terms of poverty rates and needs-based caseloads depends, at least in part, on how long the economy remains strong. Stay tuned for a broader statewide look at this next week.

Posted by: Josh Lehner | October 17, 2016

The Great Recession Didn’t Kill the Business Cycle

There is a clear distinction between the underlying trend and the cycle around the trend. Just take the 2000s as an example. The cycle component was, unfortunately, very evident given the two recessions, even as the underlying economic trends were flat or down for the most part. While I don’t believe most observers really think the business cycle is dead, I do think there is a segment that believes we experienced a really big step down in the Great Recession and haven’t or won’t see any real progress.

Our office’s view is that there is a natural ordering of events over the cycle. We generally know what portion is coming next, even if we can struggle with the pesky little issue of timing the events. The good news is that the economy really is getting to the point where improvements are being seen, or will be soon, in some deeper measures of economic well-being. The difference this time was the severity of the Great Recession and the amount of recovery needed.

A stylized ordering of the cycle goes something like the following. During a recession job losses ensue. In the early recovery job growth returns but wage growth lags as the pool of candidates greatly outnumbers available jobs. As the labor market gets tighter, wages start to rise. As employment and wages rise, household incomes for those in the middle and bottom part of the distribution also increase. Then we start to see poverty rates and needs-based caseloads begin to fall. After a period at or near full employment, the cycle starts again. Measuring progress on poverty and caseloads, to a certain extent, is about how long the economy remains at or near full employment, or how long before the next recession begins.

Of course, this stylized ordering leaves a lot of issues out. In particular these include structural changes and the trends facing different segments of the population or regions of the state. However, the larger point still stands.

On Wednesday I am presenting at the Multifamily NW’s Apartment Report breakfast where I will be highlighting some of these broader improvements seen in the Portland region. I will post a summary of that here on the blog as well.

It should be said that we are seeing these improvements first and foremost in the Portland metro area, while improvements are considerably smaller to date in the rest of the state. I think, and will detail this a bit further, that this is mostly a timing issue. Portland’s economy, like large metros across the nation, returned to growth first and has seen strong gains since 2010 or 2011. The state’s other urban areas and much of rural Oregon spent around three years at the bottom of the Great Recession. They stopped losing jobs, but saw no net growth for years. However, job gains have returned in recent years. The state’s second tier metros are now surging. Rural Oregon job growth is now back to rates seen during the housing boom. However rural Oregon has yet to regain all of its recessionary lost jobs and growth rates are lower than those in urban Oregon.

All told, the 2015 American Community Survey data brought good news. Given everything we know today, the 2016 data, when released a year from now, will show another big step in the right direction. Of course further progress remains to get back to 2007 poverty rates in much of the state, let alone 1999 or 2000 rates. That said, the economy is getting to the point where at least some of these improvements become reality and not just the next phase of the business cycle.

Posted by: Josh Lehner | October 13, 2016

Inflation and Housing

There’s no question that the cost of housing is increasing significantly faster than inflation overall. In fact, in the past year or two, inflation excluding housing has actually been zero or negative. That is, if you take housing out of the calculation, there is no inflation in the economy.

This is potentially worrisome for a few reasons however I want to focus on one in particular, based on recent conversations with the Governor’s Council of Economic Advisors. One concern is that as core inflation firms or picks up closer to the Federal Reserve’s target, the Fed will raise interest rates to head off future inflation. Ok, but what if that is entirely due to housing and not due to a general increase in inflation across the board? A sector specific shock — in this case rising housing costs and inflation — is different than a general overheating economy which is the Fed’s primary concern here. Furthermore, the housing inflation is due to a fundamental lack of supply, which is something monetary policy isn’t designed to deal with either. Such a scenario can lead to policy mistakes, or so the thinking goes.


Well, as logical as that argument seems on its face, there are a number of counterpoints and clarifications that, thankfully, make it much less of an issue. Noted fed watcher and University of Oregon professor Tim Duy helped set the record straight in our conversations.

First, look at national and not local inflation. Not all housing markets are showing robust increases and the national figures are more muted.

Second, housing is an important part of so-called sticky inflation. These items are less volatile over time, hence the term sticky. They are important in setting inflation expectations about the future. Eroding inflation expectations in recent years are a major concern, so items like housing that are rising are actually helping this situation, despite the strains on household budgets.

Third, Bill Conerly suggested we actually do the math on housing’s contribution to overall inflation. As shown below, that contribution is right in-line with historical patterns. So the issue is not so much housing inflation, at least nationally, but the fact that inflation outside of housing is so low. This is also true if you exclude food and energy. 


Finally, at a more fundamental level, you should really be looking at PCE instead of CPI. The simple reason is that the Fed targets PCE and the composition of the basket is different. For example, Housing and Utilities account for 18% of actual Personal Consumption Expenditures in 2014 and 2015. Conversely, housing in the CPI basket is 42% nationally, more like 44% locally. Housing’s 42% can be broken down into Shelter (33%), Fuels and Utilities (5%) and Furnishings/Operations (4%). 

National housing trends as seen in the actual consumption data do show milder trends than strictly looking at rental increases in popular, coastal metros. That said, it’s not that those rental increases are not a problem, they clearly are. Rather it is much less a concern for monetary policy makers.

Posted by: Josh Lehner | October 11, 2016

More on Household Income Growth

Mea culpa. A couple weeks ago when discussing the latest median household income trends, I used the wrong inflation adjustment, thus skewing the 2007 to 2015 comparison. It made the data look worse than it was, even as the big picture patterns and comparisons remain the same. The 2014 to 2015 adjustment was correct. I have updated the previous post with corrected data. My apologies.

In the meantime, to better show the trends in household income at various parts of the income distribution, I have created the graph below. Don’t worry, the inflation-adjustments are correct here.

Anyway, the story remains the same. The highest-income households have seen the best trends in the past decade. The typical household’s income is now nearly back to pre-Great Recession levels, although still below after accounting for inflation. Lastly, since bottoming out in 2012, income gains at the bottom of the income spectrum have actually been the strongest.

These improvements are all about the strong labor market. Middle- and lower-income households only have wage income and the safety net, more or less. Thus when the economy gets tights — more people have jobs and wages rise — those in the middle and lower half see better income gains. A strong economy at or at least nearing full employment has different implications for the outlook as discussed before. Lastly, it should be noted that higher income households have a wider variety of income sources, including capital gains, rental income, dividends and the like which have generally performed better; thus their better trends in recent years (and decades).


Posted by: Josh Lehner | September 27, 2016

Peak Renter In Real Life

A year ago our office asked “Is 2015 Peak Renter?” We laid out a straightforward case examining the three main underlying drivers for the shift into rentership over the past decade: household finances, demographics, and preferences and tastes. We know the pendulum has swung all the way toward rentership. The question is not whether or not it will swing back toward ownership, it will. The question is when will this happen. Our office’s position was that it would happen probably sooner than the conventional wisdom suggested. The reason was household finances and demographics are now working in favor of ownership.

Overall that piece got a lot of attention. Bill McBride of Calculated Risk even talked about it on one of his Bloomberg TV appearances (thanks Bill!). I also think it is fair to say that our work was greeted with a healthy amount of skepticism as it does when we discuss it in our presentations.

Well, the 2015 ACS data was just released and we got an update on the question of ownership vs rentership. As Jed Kolko notes in his great summary of the ACS data, rentership ticked up 0.1 percentage points nationwide last year. So not yet peak renter across the entire U.S. and the HVC data differ from the ACS data as well. That said, the U.S. changes were minimal, suggesting the shift into rentership has slowed considerably; possibly an indication that peak renter is near.

Locally, however, peak renter is already here. Statewide the homeownership rate ticked up from 60.7% in 2014 to 61.1% in 2015. In the Portland metro region the increase was even bigger at 1.6 percentage points.


Yes, it is just one year. One data point does not make a trend. It is possible that some of this increase will be given back in the 2016 data. However decomposing the ownership gains by age reveals an interesting pattern. This is what peak renter looks like in real life (IRL). The 20- and 30-somethings are seeing the largest ownership increases. The vast majority of these households rent (~80% renter in early 20s, dropping to 50/50 in mid-30s) but on the margin ownership is coming back somewhat.


Now, how pervasive are these gains across the nation? Is this just an anomaly or can we chalk it up to noisy data? In short, no or at least probably not. Looking at the nation’s 100 largest MSAs — the same 100 that we studied in the Housing Trilemma — shows that half of them saw their ownership rate increase in 2015. This is considerably more than the 20 or so that saw yearly gains during the Great Recession and its aftermath. Those 20 are probably more about noisy data than any real return toward ownership.top100ownership

When decomposing ownership gains by broad age group, the largest metros show a similar pattern in that more are seeing increases among the younger cohorts. Some of this is due to composition effects where household formation among the young is suppressed as more are living at home. This influences the potential young renter households more.


However, the issue is not the magnitude of the changes but the direction.  As we argued last year, two of the three main reasons for the increase in rentership — household finances and demographics — are swinging back in favor of ownership in recent years. And expected to continue to do so. The third main reason — preferences and tastes — remains more of an open question.

As for the outlook, our office expects ownership to continue to increase somewhat in the coming decade. Just how far is uncertain. We are not likely headed back to 2007 ownership rates, nor should we. The large transaction costs associated with buying a home and the lack of flexibility is not for everyone or for every household. However the pendulum had swung all the way in the other direction in the past decade. More balance is likely to be seen moving forward.

It is also worth noting that in growing metropolitan areas like Portland, the overall number of renters and owners will both increase. There will be more of everybody. The challenge is ensuring the housing supply keeps pace with demand, which has not happened in recent years. Affordability has eroded considerably as a result.

Posted by: Josh Lehner | September 22, 2016

Household Growth by Income (and Housing)

Our office talks a lot about migration and population growth. It is one of the key reasons Oregon outperforms during economic expansions. However, the changing composition of the economy and households underneath the topline has a big impact. This is particularly true when it comes to the housing market in recent years. A year ago, in work I was doing ahead of a presentation for Multifamily Northwest, I detailed how job polarization was also impacting household growth by income in Portland. At what incomes levels were we seeing the growth? However this not just a Portland story, it is seen throughout much of the state.

Yesterday, Mark and I gave a housing overview presentation to the House Revenue Committee. We tried to hit all the high notes and the big picture supply and demand trends we are seeing. However on the household growth by income level chart there was a bit of confusion and I want to clarify what I was trying to say. This is because I think this growth is one of the key aspects underlying the current housing market. This is important and I failed to properly communicate what I wanted to yesterday.

First I should have started with a graph that looks at the number of households by income level. The three largest groups of households in Portland are the three income groups below $75,000. This, of course, make sense given median household income is $64,850 per the latest ACS data. As such, 58% of households in Portland earn $75,000 or less per year.


However, the story is different when you look at the net change in the number of households by income level in the past decade. On net, when you compare the composition of households in 2015 to the composition of households in 2007, all of the growth has been at the top end of the income spectrum. So even as these higher income households represent a minority of the overall population, they represent the majority of the growth.

This has big implications for housing. These higher income folks can afford higher housing costs. The rough affordability numbers listed below use the ballpark rule of thumb that housing should not exceed 30% of income, even as that is an imperfect measure. A household earning $100,000 per year can afford $2-3,000 per month in rent or to purchase a $500,000 home. Most Portlanders and Oregonians cannot of course. So when the conversation turns to the fact that we are only seeing new construction as these luxury price points, it is not just because margins for developers are better at higher prices. They are of course. However they are also responding to a fundamental shift in where the demand for housing is coming from, namely, from the higher income households. That does not mean we do not need more housing at lower price points; we desperately due. And a large part of our office’s focus yesterday was trying to look at the broken supply side of the housing market. We are also working on drafting up some of our comments along those lines.


Below is a similar chart but looking at the Eugene MSA (Lane County). The same general pattern is seen and this is actually what we showed yesterday, not the Portland data. Notice that the housing affordability numbers do not change. That is because it is based on the 30% rule of thumb.eugenehhincchange0715

Like Mark said yesterday, housing affordability is not just an urban problem today and it certainly is not just a Bend, Hood River and Portland story. Rural affordability has many challenges as well. Representative Bentz asked us to follow-up on some of our rural comments and charts and we will do so in the near future.


Posted by: Josh Lehner | September 20, 2016

Household Income Growth by Quintiles

Noted in yesterday’s look at how median household incomes are finally rising was that, at least nationally, the lowest incomes increased by the largest amount. That was based on the CPS data available for the U.S. In the ACS data, which we have at the local level you can do something similar as well, which brings us to the latest edition of the Graph of the Week.

In 2015, incomes for Oregonians in the bottom quintile — the 20% of households making approximately less than $23,000 per year — increased by the largest percent. This amounted to nearly a 7 percent increase after accounting for inflation. Gains were relatively even for the rest of the income distribution. In fact this pattern of strong gains at the bottom end, plus relatively even gains for everyone else is seen in Oregon data since the depths of the recession, or so far in recovery. These gains are predominantly due to the improving labor market where wage gains are increasing as the market becomes tight.

UPDATE: The graph below and percentages listed were incorrectly calculated in the original post. The issue was using the wrong year for inflation-adjustment. What follows is a corrected version. The overall story remains the same, although the exact details do differ. Apologies for the error.

However, when looking at income changes from before the Great Recession through today, there is a clear pattern. Those at the highest income levels have seen the best performance. Inflation-adjusted incomes for the top 20% are now nearly 7 percent above their pre-Great Recession peaks. The top 5% have fared even better with gains of 8.1%. On the other hand, incomes for the bottom 20%, even after the big jump in 2015, remain 6.5% below where they were in 2007.


Our office is still unpacking and dissecting the latest data. Stay tuned for more updates in the coming weeks.

Posted by: Josh Lehner | September 19, 2016

Oregon Median Household Income Rises (Finally)

The big economic news last week wasn’t our office’s forecast, but the big jump in median household incomes nationwide. Two separate reports showed two different numbers: the Current Population Survey gains were 5.2% and the American Community Survey gains were 3.8%. The big takeaway isn’t the differences but the strong growth seen in both. The recovery is finally translating into income gains for the majority of households. In fact, per the CPS data, the largest percentage increases in incomes were seen at the lower end of the income spectrum. This coincided with a relatively large decline in the poverty rate as well, as would be expected. Overall certainly good news and a welcome reprieve from lackluster to declining incomes in recent years. What is driving the results? I’ll let Jed Kolko, writing at Calculated Risk, explain:

Most of the jump in median household income, therefore, appears to be rising earnings, with rising employment playing an important supporting role. The labor market improved for workers on both of these fronts: the rise in median household income is indeed good news.

ACS data was also released for state and local jurisdictions* and the Oregon data looks even better. From 2014 to 2015, Oregon’s median household income grew by 6%, the third fastest rate in the nation behind Montana and Tennessee. The poverty rate dropped by more than a percentage point as well. However, from 2007 to 2015 Oregon’s median household income gains rank just 32nd best among the states and DC. And even with the poverty improvements, the state is only about halfway back down to pre-Great Recession poverty rates. Still, progress is being made and the gains are seen throughout most of the state as well.

Below is an update to our office’s graph comparing median household incomes for Oregon and the U.S. The 2015 gains mean Oregon’s median household income is 2.9% below the U.S. median, which is more in-line with the typical year where our incomes trail the nation by 2-3%. In inflation-adjusted terms, Oregon’s income in 2015 is about 1.5% below 2007 levels and 2% below 1999 levels, at least when using the PCE deflator. Using different deflators or measures of inflation will change those percentages, sometimes drastically so.


In recent years our office has been hitting on the major theme that robust job growth in Oregon has translated into solid wage gains as well. This is not the case nationally. It terms of the data, it was more a matter of time before the stronger Oregon labor market translated into median household income growth and lower poverty rates. 2015 is a start. Given the ongoing economic strength, 2016 is shaping up to be even better. A key question for the outlook is whether or not the economy can get fully healthy before the next recession. This did not happen in the housing boom. The cumulative progress made since 2009 or 2010 has actually been very significant even if year to year changes are muted. Full employment is now within sight, although not here yet. Broader gains across the board are now being seen, however incomplete they may be thus far.

* All of the 2015 ACS published tables are now available. Our office is working on dissecting the latest data and updating our records. The good news we have new data (yay!) the bad news is we have new data (have to update everything…). The microdata won’t be available for another few months. Then the real work begins when we can crunch the underlying sample data.

Posted by: Josh Lehner | September 14, 2016

Oregon Economic and Revenue Forecast, September 2016

This morning the Oregon Office of Economic Analysis released the latest quarterly economic and revenue forecast. For the full document, slides and forecast data please see our main website. Below is the forecast’s Executive Summary.

The economic expansion continues. However a few mounting concerns muddy the outlook with risks clearly tilted toward the downside. The good news is the nation continues to muddle through various headwinds with job growth strong enough to bring down headline unemployment and make some progress on underemployment. The strengthening labor market, better wage gains and strong household balance sheets continue to support the U.S. consumer, the only real economic bright spot in the past eighteen months.

Even so, there are a number of worrisome trends that have emerged in the data that should give forecasters pause. First, manufacturing and industrial production continues to be weak. Second, personal income growth nationwide is slowing. Third, due to slowing tax revenues, an increasing number of state revenue forecasters are missing their forecasts. Fourth, the dearth of new business investment is weighing on growth. All told, even with these concerning trends, the baseline outlook still calls for the expansion to continue. And while expansions do not die of old age – they die due to mistakes – the economy is clearly closer to the next recession than not.

While some U.S. data is slowing, Oregon’s expansion continues to see full-throttle rates of growth. Oregon is outpacing the typical state by a considerable margin today for both job and income gains. This growth differential largely comes from the state’s underlying fundamentals like its industrial structure and strong in-migration flows. Both of these trends have long-lasting impacts on the Oregon economy and help drive the state’s more volatile swings over the business cycle.


Oregon’s General Fund revenue outlook remains stable. Personal income tax collections continue to expand at a healthy pace, keeping revenues in line with what was expected when the budget was drafted. Oregon’s General Fund revenues are currently expected to end the biennium within 0.1% of the Close of Session forecast.

Personal income tax collections continue to reflect Oregon’s strong underlying labor market.  Withholdings out of paychecks expanded at an 8% rate during fiscal year 2016.  As such, state revenue growth in Oregon remains among the strongest in the U.S.  State revenue growth would have been even more rapid in recent months if not for the payout of the personal income tax kicker generated during the 2013-15 biennium. The vast majority of kicker payments have now been made, and will no longer weigh on overall collections.


In contrast to the healthy growth seen in personal income tax collections, corporate tax collections have been falling sharply in recent months. Nationwide, corporate profits have taken a step back, largely due to rapid appreciation of the U.S. dollar and struggles among energy firms and other commodity producers.  Even so, profits and corporate tax collections remain large relative to historical norms. Given the expectation that collections would return to earth, revenue declines were built into the forecast, leaving the outlook very close to the Close of Session forecast for now.  Declines are expected to continue through the current fiscal year, further reducing annual revenues by around $50 million.

In addition to healthy General Fund revenue growth, Oregon Lottery sales and Estate taxes have been very strong as well. Recent collections have consistently come in above expectations.

Revenue growth in Oregon and other states will face considerable downward pressure over the 10-year extended forecast horizon.  As the baby boom population cohort works less and spends less, traditional state tax instruments such as personal income taxes and general sales taxes will become less effective, and revenue growth will fail to match the pace seen in the past.

See our full website for all the forecast details. Our presentation slides for the forecast release to the Legislature are below.

Posted by: Josh Lehner | September 12, 2016

Careful with the Data, Job Polarization and Housing Math editions

Ahead of Wednesday’s forecast release, two quick notes on items that caught my eye recently.

First, as our office has written quite a bit about in recent years, job polarization continues to shape the economy and labor market. That is, jobs are increasingly concentrated at the low- and high-ends of the wage spectrum with shrinking opportunities in middle-wage occupations. The Federal Reserve Bank of New York, who pioneered the local and regional polarization work our office’s uses, has a very useful update they are using in their presentations. One key point when it comes to polarization is that these jobs do not decline forever, they do grow in absolute terms during economic expansions. In fact, as the NY Fed’s graph shows, middle-wage jobs increased in absolute terms faster than both low- and high-wage jobs in recent years. This is good news!


However this does not mean that job polarization does not exist. Middle-wage jobs should increase more in absolute terms given that they represent a larger share of all jobs. The problem with polarization is that middle-wage jobs decline the most in recessions and do not come back all the way during expansions. Thus as a share of the economy, middle-wage jobs are shrinking. This can be seen in the second graph which uses the exact same data and same categories but looks at growth rates instead. Clearly middle-wage jobs, while accelerating some, are still growing slower than both low- and high-wage jobs.


The point here is that it is important to look at both levels and rates. One can potentially be mislead when looking at just one part. The NY Fed did not try to mislead and I am not accusing them of that. Rather, I just wanted to flesh out the story a bit further. I also teach a data analysis and presentation module at Willamette University and this provides a good, new example to bring to class. There is considerably more art and less science than many think when it comes to data analysis and presentation. You need to understand the whole picture to know which parts you want to emphasize in your work. And emphasizing the fact that middle-wage jobs do not decline forever and are even doing much better in recent years is certainly important to do. That point does tend to get lost in the polarization discussion.

Item two is more of a funny data note. Coldwell Banker Real Estate crunched some numbers on the average list price for 4 bedroom, 2 bathroom homes across the country. As the Portland Tribune writes, in Oregon the highest list price is in Lake Oswego and the lowest average list price is in Klamath Falls. From this one can conclude that LO is the least affordable and K Falls is the most affordable place in the state, with the caveat you are trying to buy a 4 bedroom, 2 bathroom home today. Just to check, I wanted to see how these list prices compared with family incomes in the area. It turns out that the list price to family income ratio is nearly identical. While there are a variety of reasons that homes are more expensive in Lake Oswego and less so in Klamath Falls, one of the reasons is income. Just an interesting note on housing and affordability, or lack thereof.



Older Posts »